Centre de Pedologie Biologique, CMRS, \landoeuvPe-les-Nanc1Q, 54500, Frente and QRSTQM, Dakar, Senegal Present address: Soil Department, Pdoscow State University, Moscow 117234, USSR

1. INTRODUCTION e

It is well known that, even in the absence of any fertilher or manure, flooded rice (Oriza sariva) can grow in the same plot for centuries without any. depletion of the yield. The maintenance of soil fertility has been assumed to

1-3 result mainly from asymbiotic N2 fixation by soil or water microur~tnisms. Three categories of microorganisms may be involved: ( i ) pee microorganisms, blue-green algae and photosynthetic bacteria ?- blue-peen algae, such as Anabaena associated with various Azolla trpecies. and ( 3 ) heterotrophic bacferia associated with algae or with the rice root system. Since excellent reviews have already been presented of the recent work on ecological aspects of N, fixation by the two former groups of microorganisms, this note conceñtrates on N2 fixation by the free-living heterotrophic bacteria of the rice rhizosphere. Bur aim is to provide a survey of the latest findings on: (1 ) the application of the C2H2C?H4 assa? tu complex rice-soil systems: (2) the use of gnotobiotic systems (mde l s y t m i * I (3) factors affecting asymbiotic N, fixation in the rice rhizosphere; (4) tkld estimations of N2 [C2H2] fixation: and (5) the populations of diåzotrophs ln the rice rhiLosphere.

2. APPLICATION OF THE C2H -@?Hq ASSAY TO COMPLEX RICE-SOIL S?STh4S

The various uses of the C2H2-C2H, assay have been conipre!iui~sively reviewed." However, some specific difficulties may anse due to $3 mpling and incubation chambers, diffusion of C2H2 and C2H4 thrutigli Pic' plant-soil system, ' duration of incubation and illumination during incubation, and extrapolation of N2 fixation data in time and in space.

2.1. Sampling and heubation Chambers

Three types of sampling and incubating procedures (for a n t plant-soil system) have been developed for field C2H2-C,-&B4 ~ i ~ s t , ~ 76 IfDtdkd. excision or

'Present address: Soil Department, Moscow S t a t e ~ j p i ~ ~ ~ , ~ ~ ~ ~ ~ 1 7 2 ~ ~ ~ ~ 5 ~ ~ ~ ~ ~

j2 $ /\@Bjr 1pf;;i

- j p * c

rhizosphere soil collection, followed by incubation in ,afferent kinds of closed chambers;"*" (2) extraction of soil cores including the plants to be studied, which are incubated in air-ti&t and (3) field techniques employ- ing internal standards.

When plants are grown in jars or tubes under controlled conditions in the laboratory, the entire &at++oil system is usually incubated with C2H2 in different kinds of chattikic (f'tg. 1). Whatever type of plant-soil system is studied, the saniplrnp 9': c x d u r r should disturb flic system as little as possible. The use of h i r i ) h g c unduiturhed cores13 o r the ftcld technique employing internal standardst4 is. thrreiwt. rtcummendcd w~bmmer possible.

14

2.2. Diffusion of C?Hz ;mil C through the Plant-Soil.6ystem of Rice

The rice plant transports O2 from the atmosphcrs - % the i o u t b ff id the

(ref. 17) and C2H2 (ref. 18), are present in the gas phase r j t A rice-sod system, they would be expected to diffuse readily to tlie rhizosphere through the plant. This assumption was only partly supported by the results of a prelimi- nary experiment with C2H2 (Fig. 2). Rice (IR8)-soil (Boundoum) systems developed in the device Dl9 (Fig. 1) were submitted to two different kinds of incubation. C2H2 was injected into either upper compartment U (6 replicates) or lower compartment L (5 replicates). Fig. 2 shows that slightly less C2H4 (14.8 i 2 nanomdeswxding. hr) was produced in the lower compartment WheH ,Ct, EIS@ L:~~s*v m : t 1 C I ? upper (impartment than when i t was injected tofu *.L.. . m 1 A II r .:n&s!seedLing. lit). Moicliver. c H &LCrjm&&? 7 t a s. I I I I I C ~ I smaller than that in

L I ~ B C ~ ~ X P ~ a poor diffu- the lower &ïm*@ IL.

sion of C2H4 from w t r ~ t -11- 1 - i 7s .**il,. I .( iØ? rintc~ which gases diffuse matly through me), v t ~ AT%.: m $ t , .t wP7 art\ khis pathway to ensure a satisfactory diffusion ni 1 +$: 3 1 c2&& ~ ~ t m dta gas phase to rhizosphere and vice v e m , especially when large cores or field devices14 are operated. The la , which occurs during %H C H assays (reported for the

degree this hindrance in gas diffusion.

2.3. Duration of Incubation

rhizosphere soil via tissue gas spaces. 15J6 Thus, \viler1 *tt\cr gases, 6 g N*

- . _

2 4

rice-soil system" as well as other complex systems 2 ?l ),miat reflect to some .

c -

Estimations of N2[C2H2] fixation rates based on long-term incubations are higher than those based on short-term incubations (Fig. 3). 'Thus firation rates, estimated at the beginning of the light period, were only 0.5 micromoles C2H4/g dry ro0t.h when based on 1 hr incubations, but were 3 4 micromoles C2H4/g dry ro0t.h when based on 8 hr incubations. Corrapmdlnp rates, estimated at the beginning of the dark period, were 0.2 and 1.0 pertpectivtly.

B c D

Figure 1. Incubation devices for measuring N2[C2H,] fixation in rice-soil systems: A. classical Serum bottle; B. double test-tube device; t. beaker with incubation canopy made of polyethylene bag; D. device allowing C H2 separate injection into upper (shoot) compart- ment and lower (root) compartment.19 1. Rice seedling; 2. Soil; 3. Water; 4. Rubber cap; 5 . Serum bottle (500 ml); 6. Pyrex test tube (14 x 140 mm); 7. Rubber attaclunent; 8. Beaker (150 ml); 9. Polyethylene bag (50 x 300 mm); 10. Sampling port; 11. Paraffin oil; 12. Red sand (for isolating algae from light); 13. Pyrex tube with one side arm (14 x 100 mm); 14.Pyrex tube (14 x 200 mm); 15. Pyrex tube with two side arms (18 x 180 mm); 16. Vigreux points; 17. Vacutainer stoppers; 18. Paraffin oil seal (for isolating algae from C2H2); 19. Vacuum tubing.

High N2 [C2%] fixation rates induced by long-term incubations may be attributed to either a direct effect of C2H2 on rhizosphere bacteria21 or an indirect effect through a stimulation of exudation. Such stimulation could possibly result from pathologic reactions of the plant to %HZ 81 C2H4, or,

1 o 5 IO

1,“ (I,

7-

1lY .”. 1

Figun. 3,. Time uour\e of C H -C*% reduction in a rice(lRS)-wil s> stem. uung incubation device D. Fipure I . U. $3, kjt$teid into upper conrportmenr: U U . C,t$ produced in upper compartment; UL,% k produced io lower compartment: UU’. C- ti remaining in upper compartment; UL’, Ed2 remaining in .Ipwer compartment. L. C;H, injected inro lower cotnpartmetir: LU. C 74 produced in upper compartment; LL, c,1f4 produced in lower compartment; LU’, ¿!,il2 remaining in upper compartment; LG‘, C ~ H ~ remaining in lower compartment.

’ 2

more likely, from an increase of photosynthesis due to the accumulation of CO., in the incubation chamber through soil respiration. Thus long-term incGbations should not be used, except when tfiere is no other way to overcome inadequate gas diffusion into the soil.

2.4. Illumination during3ncubation

During incubation plant-soil systems should be illuminated in conditions closely simulating those in the field because the activit of N2-furing bacteria in the rhizosphere is influenced by photosynthesis. 1 3 a 2 % Experimental results summarized in Fig. 2 show that illuminating a rice-soil system previously placed in the dark promoted N2[C2H2]-fixing activity. The problem of diurnal variations of N2 [C2H2] -fixing activity induced by day-night sequence will be discussed later.

514

" r LIGHl

r 12 i 4

Time (h)

Figure 3. Estimations of NZ[C2H2]. fixation of a dœ(IRI)-soil system based on short-term or long-term incubations. e---e diumal variations of N (C2H2] -fixing activity based on short-term incubations, limits show standard errors of means; - cumulated curve of CzH2C2H4 reduction over long-term (up to 12 hr) incubations. For long term incuba- tions. then were only 2 assays, thus no diurnal variation curve could be drawn.

2.5. Extrapolation of N2 Fixation Data in Timb and in Space

The conversion of N2 fixation data obtained from necessarily limited sampling units (excised roots, cores, field incubation devices) to fairly large areas involves sampling on a distribution pattern W i g into account variations in soil char acte ris tic^?^ However, such spatial variations are assumed not to be of much consequence within most paddy fields: More important are diumal and seasonal priations in N2 fixation activity. These changes re uire that sampling be performed at regdar intervals thfoughout 24-hr periodf2 and at many times throughout the yew. Only integrated values based on data obtained in such a way can be considered valid.

3. USE OF GNOTOBIOTIC SYSTEMS

Ecological studies on N2 fixation in the rice rhizosphere require the use of gnotobiotic systems comprised of sterile rice seedlings and a known popula- tion of diazotrophs. A rice (IR8)-Beijerinckia model system was found to fH N2 as actively as did complex rice-soil systems studied in the laboratory (Table l), the activity averaging 1,000 nanomoleslg dry root.hr. In the field, the extremely high activity observed (30,000 nanomoleslg dry r0ot.h) was attributed mostly to blue green algae. Gnotobiotic systems such as ours, which exclude algae, confirm the postulated high potentiality for N2 fHation of ricediazotrophic bacteria associations.

4. FACTORS AFFECTING ASYMBIOTIC N? FIXATION IN THE RICE RHIZOSPHERE

As the rhizosphere ir by no mems d n isolated compaiftrient nf the plani-roil system, microbial activity in general and asymbiotic N,4ixuig activity in the . rice rhizosphere in particular k bependent not only upun wit factors but also upon atmospheric parameters and, moreover, upon the plant itself. Table 2 summarizes the different factots that have been studied to date in the rice-soil system.

4.1, Climatic Factors

The effects of illumination have been investigated in ‘growth chambers and in the field. In growth chambers, N [C2H2] fixation of rice seedlings (20-27

rootbr) than at 30,000 lux (2,000 nanomoles/g dry r0ot.hr)2~ Fig. 2 shows that, 1-2 hr after illuminating a rice-soil system,. nitrogenase activity was significantly stimulated. In the field (Fig. 4), diumal variations induad by the day-night sequence were noted?2 in growth chambers, diurnal variation curves were similar, except for the absolute maximum activity, which was 1/10 that

days)-soil systems was 10 times T ower at 3,000 lux (200 nanomoleslg dry

616

~~ r I c

Table 1. N [C H 1 -fixing activity of various rice-soil systems and of a gnotobiotic system. 2 2 2

Incubation

Device Timefir)

Field experiment Lamto soil, Ivory Coast cd 3 (Taichung native 1 rice)

Laboratory experiments (i). Rice (IRB)-soil systems

Djibelor soil, Senegal A’) 24 Djibelor soil, Senegal B(b) 3 Camargue soil, France C@) 3 kundoum soil, Senegal D’I 3

NZ[C2H 1 -fixing activity Presence (nanomo?es C2H4/g dry root.hr)

of algae Day Night

+ 30,000 2,000

( I I ) . (hoiobiotic system Rtu.i1RS)-Brijerinckkr sp. (4 3 O 1,000

14 (a) Field devicc ~ C % L rihpd prrvcously. (b) Cf. Fig. 1. (c) Algae were present. but their activity was stopped by applir.iiron of sand Arid paraffin oil at the soil surface. (d) Not described here. 2

4

3,800 200 1 O0 600

i - ,

Table 2. Factors khdwn to affect N2 asymbiotic fiaation in the rice rhizosphere.

Atmosphere : light; CO2; air humidity.

: N2 and O2 diffusion through the plant; [ Plant Ecosystem compar~ents exudates.

Rhizosphere soil : water content; soil type; dissolved gases (N2, 02); horgwic and organic N: potassium.

o 24' Hour6

Figure 4. Diurnal variations of N2(C2H2)-fìxing activity measured in a paddy field at h t o (Ivory Coast). Limits show standard errors of means.

618

of the in situ pig. 5). N2[C2%] fixation in some plant-soil systems was shown to be very sensitive to the water stress induced by a decrease in air humidity?2 This effect is supposedly attributable to stomatal closure; in the case of flooded rice fields, however, this climatic parameter should not be a major lhiting factpr:

c i

æ

* O O

1.

1.

L lGhT

\

2 0

0 7 \

I u N

O i 2 24

Figure 5. Diurnal variations of N [C H ]-fixing activity of two rice(IRI)-soii systems growth chambers: 0 4 with light duration of 14 lu, incubation device as Figure 1.B, Boundoum soil;@,+ with light duration of 15 hr. incubation shown in Figure 1 .C, Camargue soil.

2 2 2

DARK

T i e (h)

placed in shown in device as

659

I

I Atmospheric "02 c m t n t i o n signnificimtly affects. N2 fMtion in vmbiotic systemsz6 and ir uwmd to play a similar role in rice-soil systems. Its sigaificance in the field k, as yet, unknown, but preliminazy laboratory investigations show that either CO2 depletion or CO2 accumulation (vide supm) during C2H2 incub&an may explain some anomalous results. Thus, N2[3%]-fixing activity d o gnotobiotic rice (IR8)-Beije~ackia system was only O0 nanomoles/g dry rooth when growing in a chamber where the gas phase was not nplenirhcd, but exceeded 1,500 nanomoles/g dry root.hr four days after a CO2 injection nisin8 the CO2 content of the gas phase up to 5,000 ppm (1. F u c d h m a d , personal communication).

4.2. Edaphic Factors

Waterlogging is known to favor mbiotic N2 fixation in an s y ~ t e m . 2 ~ The ricesoil system is no exception?'Yoshida's investigations 25 $9 illustrate this fact and, in addition, stress the major role of the rhizosphere as N2[C2H2] fmation is consistently much higher in planted than in unplanted plots (Fig.

Dissolved or adsorbed gases in paddy soils are apparently at least partially provided by the rice itself (vide infia).

Combined forms of N affect N2 fixation in complex systems and pure cultures similarly. Yoshida er al.30 noted that 160 ppm of added combined N completely inhibited N2[C2H2] fixation in a rice-soil system. Similar results were found by Balandreau eral.,l8when 100 ppm of inorganic N inhibited 90 % of the N2[C2H2]-fixing activity. If the. absorption rate of inorganic N by rice is higher than the inorganic N influx in the rhizosphere, the rice plant could indirectly contribute to the derepression of N 2 m in the system. Soil organic N did not appear to restrict N2[C2Hz] f i t i a n , as three ricesoil systems with increrdtg organic N content (0.1 5 1 ,0257 I 0.3oQ 7%) exhibited

6).

21-fixing activity of 10. 20. 38 ~~lizmokr C2H& dry

Beside m i n e d M, the status of other nuttrerrts. namely P, Mo and K, may also affect N flxation by alteshg plant exudation. Trolldenier3' found that insufficient 2 nutrition gave rise to larger excretion of organic com- pounds by rice roots, causing higher bacterial activity. This result suggests that N2 fixation could be indirectly stimulated by K deficiency.

4.3. The Plant Factor

A most prominent, though poorly understood, factor controlling Npse activ- ity in a plant-soil system is the plant Yoshida and AnCojas? repotted that N2[C2%] fixation of rice systems involving a Pera variety was four- times higher than that of a system with an IR8 variety. The effect m y

.-

1 3 5 7 9 1 1 13 15 17

Weeks after transplantfng

Figure 6. Variations of N [C2%]-fixing activity in planted (rice IR20) and lanted upland and submerged soilsaunng 1971 dry season at Los Baños (Philippines). 3%

be attributed to differences in: (1) exudate compo~it ion?~ (2) amount of exudation; and (3) diffusion of gases (N2, 02) from the plant shoots to the rhizosphere. 29

5 . FIELD MEASUREMENTS OF N2 [ C2H2] ,FIXATION

Estimations based both on N balance and CzH2C2H4 assays demonstrated the high potential of flooded paddy fields for N2 fixation. A N balance of a paddy field in the IRR1 experimental station suggested that N2 input from diazotrophic organisms approximated 70 kg N2/ha.yr as the N level of the soil remained constant (Table 3). C2H -CzH4 assays performed in Japan, Philippines and the Ivory Coast gave dgures of the same order of magnitude (Table 4). Other graminae-soil systems located in temperate (BrachVpodium ramosum, Dactylis glomerata, Lolium perenne, Zea m . y s ) or in tropical (Loudetia simplex, Hyppwhenia sp., Panicum maximum) conditions appeared'

Tabla 3. Nitrogen b a k e in IRR1 experimental farm at Los B a h r , Philippines. 25

N-Sourcc

Rice grain Rice straw Rain water

Irrigation water

50.0 25.0 - -

- 2.3

3.8

to be much less efficient (Table 4). Paddy fields often deserve &e reputation of highly effective N2-fixing ecosystems; however, some poor NTfixing paddy fields also exist. Such paddy fields were recently studied in Morocco, where N2[C2H2]-fixing activity in the soil was low, even though the activity was associated with planktonic Gfoeotrichia species36 These data suggest that bacterial N, fixation in the rice rhizosphere was of limited importance there.

6. DIAZOTROPHIC BACTERIA OF THE RICE RHIZOSPHERE

&

Preliminary investigations showed that aerobic diazotrophs may predominate in the rice rhizosphere (Table 51, where a relative high p02 exists. 15,37

Diazotrophic non-photosynthetic bacteria isolated from rice rhizospheres be- long to the following genera: Pseudomonas, Azotomonas, Azotobacter;

many soils, even neutral ones. Azorobacier was recently found to play only a trivial role compared with other diazotrophs. Although the infrequent occur- rence of Beijerinckìa in temperate soils has been reported, it may have been overlooked elsewhere because of the lack of the sensitive C2H2-C2H4 pro- cedure now in use in many laboratories.18 Cultures of diazotroph associations grown in Feodorov-Kalininskaya medium* were found to reduce C H more actively than did monospecific cultures (Fig. 7). However, the stabdity and efficiency of such associations in the rice rhizosphere have not yet been tested.

A strain of Beijerìnckia sp. isolated from a rice rhizosphere in Camargue (France) profusely colonized rice roots in gnotobiotic systems (Figs. 8 and 9). The observed pattem of colonization was not very different from that of Rhizobium on Trifolium r0ots.4~ The colonizing ability of Beverinckia strains therefore confirms previous results concerning the establishment of Beijerinckìa in the rice rhizosphere?8

Be@eriticCiia, Flavobacterium, AFthrdbacteF, Bacillus, ClÖstridium. 18,3844 ln

2 . 2

*Feodorov-käliiinskaya culture K2€IP04, 1.74 g; KHzpO , 0.91 g; MgS04:7H O, 0.3 g; CaCl .6H O, 0.1 g; NaCl, 0.5 g; Fea3. 6H20, O& g; trace elments &tion, 1 ml; ye& ex&act, 10 mg (expressed as N content); carbon murce, 10 g; dbtilled wmr, 1,000 ml.

Table 4. Field measurements of N2[C2H2] fixation in paddy fields and some gramineous eco- systems.

Ecosystems

Paddy fields (Oriza salivo) - unplanted

planted

planted

Gramineous ecosystems

flrachypodium ramosum mod Doctylis glomerata sod (cv. Floreal)

Lolium perenne sod (cv. Viktoria)

Zen mays (INRA F7 x F2)

Hvporrhenia sp. savanna

Ponicum maximum sod

' Loudetia sitnplex savanna

Location

- Los Bafios, Philippines

Los Baños, Philippines

Lamto, Ivory Coast

Montpellier, France

Nancy, France

Nancy, France

Nancy, France

Lamto, Ivory Coast

Lamto, Ivory Coast

tamto, Ivory Coast

N,[C2H2J - fixed

10-200 kg N/kg dry eil.day 42.5 ' kg N/ha/dry ,æason 79.8 kg N/ha/dry season

72 kg N/ha.yr

2-1 1 kg N/ha.yr

1.5 kg N/ha.yr

12 kg N/ha.yr

9 kg N/ha.yr

4 kg N/ha.yr

.-

Refermas

34

25

25

22

this p a w

this paper

this paper

. 14

35 this paper

22

Table 5. MF" counts (hbacteria per g of dry soil) of N2-fixing bacteria in the rice rhizosphere. 18

Non-rhizosphere soil (control)

Rhizosphere soil

Rhizosphere soil + roots

Aerobic

3 200 O00

21 900 O00

61 500 O00

1.6.

1.2-

- . _E

u 0.8- -

0.4-

Anaerobic

372 O00

1 350 O00

472 600

8

O 24 4. )? 96 I 1 0 Il- (h)

O A Or 72 96 IZG I l œ [h)

I @tire 7. Time i w m e oí C li, reduction by yuïr or mixed cultures groumg I I I glucose Feodorov- ~alinmsko\ :i liquid medium - . (A) or in wirrti Feodorov-Kahiinska)s liquid medium ( B I

7. CONCLUDING REMARKS

The flooded rice-soil system appears to be generally the most active asymbi- otic N2-fixing system known to date. However, our present ecological knowl- edge suggests that some formations growing in swamps and bogs, such as Tvpha sp. or o p e r u s pupynrs stands, could be as efficient.

Sen's hypothesis was that, beside blue-green algae, bacteria were respon- sible for the increase in the N content of some flooded soils when planted to rice. 47948 This the0 was validated in 1970 in both the field and labora- tory 1892022924~25y3T44 and more recently by studies (still unpublished) on gnotobiotic systems including heterotrophic diazotropliic bacteria but no Cyanophyceae.

The results presented liere indicate that the C2H2-C2H4 assay can be used for the measurement of N2 input in paddy fields and in laboratory riœ-soil or gnotobiotic rice-diazotroph systems. However, improvements are

624

.

Figure 8. Scanning electron micrograph of the root of a rice seedling grown in liquid medium and inoculated with Beijerinckia sp. (gnotobiotic system), showing mucigel around the root cap and a Beijeritzckia sp. colony. Very few intact Beijerinckia cells are discernible, but the lipoid globules at the end of' radi cell are always visible. Gaps in the mucigel result from lyophilisation. Magnification x 1 3.000.

- . I

Figure 9. Same material as Figure 8, but the picture is related to a root zone approximately one cm behind cap. The sheath of Beijeritickia sp. cells attached to the root is embedded in an amorphous layer, presumably of bacterial origin. Magnification Y 13,000.

625

still required to overcome the various above-mentioned difficulties, especially those which impede field studies. The identity of the diazotrophic bacteria, their physiology and their ability to colonize roots, the respective role of photosynthetic and heterotrophic diazotrophs, the significance of primary ecological determinants (such as anaerobiosis, soil nitrogen forms, P, Mo and K availability), the balance between gross N2 fmation and denitrification are but a few of the problems awaiting further experimental laboratory and in situ investigations.

From an agronomic point of view, stimulation of asymbiotic N2 futation is most desirable. In some conditions, a l g a l i ~ a t i o n ~ ~ or introduction of Azolla sp. in paddy fields has been a success. Thus far, no serious attempts at bacterization of rice roots have been performed. To the best of our knowl- edge, research on the behavior in the rice rhizosphere of derepressed mutants of diazotrophs already obtained by geneticists5’ has not yet been initiated.

Another interesting possibility would be to develop organic N fertilizers com tible with biological N2 fixation, as was recently suggested by Hardy et a l . iCor legumes.

The last, and apparentiy most promising, though not exclusive approach would be to achieve stimulation of bacterial N, fixation in the rice rhizo- sphere by: (1) increasing the carbohydrate fl;x into the roots and the rhizoplane as suggested by Havelka and Hardy?-6 for legumes; (2) increasing colonization susceptibility of the roots through genetic manipulation of the

or (3) through any other means.26 Future breeding programs with rice should accommodate this approach and .sliould by no means be confined to increasing responsiveness to nitrogen fertilizers.

8. REFERENCES

I .

.2. 3. 4. 5.

6 .

7. 8. 9.

10.

11. 12.

J . H. Becking, in “Nitrogen-13 VI Soil Planr Srudit<r.” Iniernational Atomic Agency, Vienna, 1971, p. 189. A. W. Moore, Soils Fertil., 29, 113 (1966). T. Yodiida and R. R. Ancajas, Sort Sci. Soc. Amer. Proc.. ,S 156 (1971). T. S. Chopra and J . N. Dube. P h i f Soil, 35, 453 (1971). M. Kobayaslii and M. 2. Haque, in “Biological Nitrogen ,Fixortioil in Natural aird Agricultural Habitats,” E. G . hiulder and T. A. Lie, Martinus Nijhoff, The Hague, 1971, p. 443. A. Watanabe and Y. Yamamoto, in “Biological Nitrogen Fixation in Natural and Agricultural Habitats,” E. C . Mulder and T. A. Lie, Martinus Nijlioff, The Hague, 1971, p. 403. J. H. Becking, these proceedings. A. W. Moore, Bot. Rev.,%, 17 (1969). G . A. Peters, these proceedings. R. W. F. Hardy, R. C. Burns, and R. D. Holsten, Soil Biol. Biochem., A, 41 (1973). J . Dobereiner, J . M. Day, and P. J . Dart, Plant Soil, 37, 191 (1972). T. Yodiida and R. R. Ancajas, Soil Sci. Plant Nutrir., 16, 234 (1970). -

626

.)

.,

1

.i

‘ m

13.

14.

15. 16. 17. 18.

19. 20.

21. 22.

23.

24.

25.

26. 27.

28.

29. 30.

31. 32. 33. 34.

35. 36.

37. 38. 39.

40. 41. 42. 43. 44.

45.

P. J. Dart, J. M. Day, and D. Hams, in “Use of Isotopes for Study of Fertilizer UtiIisution by Legume Oops,” FAOIIAEA (Vienna) Technical Report No. 149, p. 85 (1972). J . P. Balandreau and Y. R. Dommergues, Bull. Ecol. Res. Cornm. (Stockholm), 17, 247 (1973). Y, Ishizuka, Advances in Agronomy, 23, 24 1 (1 97 1). R. J . Luxmoore, L. H. Stolzy, and J. Letey, Agron. J., 62, 329 (1970). T. Yoshida, Y. Takai, and D. C. Del Rosario, in preparation. J . P. Balandreau, G. Rinaudo,I. Fares-Hamad, and Y. R. Dommergues, in “Nitro- gen Fixation by Ree-Living Microorganisms,” W. D. P. Stewart, Ed., Cambridge University Press, Cambridge, 1975, in press. M. Raimbault, G . Rinaudo, J. L. Garcia, and M. Bourea2, in preparation. G. Rinaudo, J. P. Balandreau, and Y. ,R. Dommergues, in “Biological Nitrogen Fixation in Natural and Agricultural Habitats,” E. G. Mulder and T. A. Lie, Martinus Nijhoff, The Hague, 1971, p. 471. F. J. Bergersen, A m t . J. Biol. Sci., 2, 1015 (1970). J . P. Balandreau, C. R. Millier, and Y. R. Dornmergues. A p p l . Microbiol., 21. 662 (1 974). E. A. Paul, in “h’itrogeti Fixation by Free-Living Microorganistns.” W. D. P. Stewart, Ed., Cambridge University Press, Cambridge, 1975, in pres Y. R. Dommergues, J . P. Balandreau, C. Rinaudo, and P. Weinhard, Soil Biol. Biochem.,z, 83 (1973). T. Yoshida, in “Annual Rcpurr, !he International Rice Research Institute,” Manila, 23 (1971). V. D. Haveka and R. W. I:. H ~ r d ? , these proceedings. R. Knowles, in ‘Dinitrogen I-i.wrion,” R. W. F. Hardy, Wiley-Interscience, N.Y., 1974, in press. V. R. Rao, T. A. Kalininskaya, and Yu.M. Miller, iMikrobiolog(va (S.S.S.R.), 42, 729 (1973). T. Yoshida and R. R. Ancajas, Soil Biol. Biochem., r, 153 (1973). T. Yoshida, R. A. Roncal, and E. M. Bautista, Soil Sci. Plant Nutr., E, 117 (1 973). G. Trolldenier, PIant Soil, 38, 267 (1973). J . DÖbereiner and J. M. Day, these proceedings. M. K. Sadhu and T. M. Das, Plant Soil, 2, 541 (1971). S. Ishizawa, T. Suzuki. and M. Araragi, in ”Proc. 2nd Symposium on Nitrogen Fixation and Nitrogen Cycle,” H. Takahashi, Ed., Sendai, 1970, p. 28. J . P. Balandreau and G. Villemin, Rev. Ecol. Biol. SOI, lo, 25 (1973). J. Renaut, A. Sasson, and W. D. P. Stewart, in “Nitrogen Fixation b y Bee-Living Microorkanisms,” W. D. P. Stewart, Ed.. Cambridge University Press, Cambridge, 1975, in press. R. Aimi, Proc. O o p Sci. Soc. Japan, 2, 51 (1960). J. DÖbereiner, Pesq. Agropec. Bras.,2, 1 (1968). S. A. Z. Mahmoud and A. N. Ibrahim, Acta Agfon. Acad. Sci. Hung., 19: 71 (1970). T. Rouquerol, Ann. Agron., 13, 325 (1962). T. Rouquerol, Ann. Inst. Pasteur, 1 2 , 319 (1963). T. Rouquerol, Ann. Agron., E, 599 (1964). T. Rouquerol, Ann. Inst. Pasteur, 1 2 , 293 (1965). T. Yoshida, in “Annual Report, the International Rice Research Institute,” Manila, 47 (1970). P. J. Dart, J. Exper. B o t . , z , 163 (1971).

627 /’

46. 47. 48. 49.

i 50. 51 . . ..

1

-

T . A. Kalininskaya, Mkrobiologiya (S.S.S.R.), 2, 345 (1967). J . Sen, Ap. J. I~Ldrh, 24,229 (1929). W. I l . Willis and V, E. Green, JI., Soil Sci. Soc. Amer. PTOC., 12, 229 (1948). G . S. Venkataraman, in "Nitrogen Fixation by Ree-Living Microorganisms," W. D. P. Stewart; Ed., Cambridge University Press, Cambridge, 1975, in press. J . K . Gordon and W. J . Brill, Proc. Nat. Amd. Sci. US, e, 3501 (1972). F. B. HOU and T. A. G. Larue, these proceedinps.

6 28